TY - JOUR
T1 - Machine learning for detecting COVID-19 from cough sounds
T2 - An ensemble-based MCDM method
AU - Chowdhury, Nihad Karim
AU - Kabir, Muhammad Ashad
AU - Rahman, Md Muhtadir
AU - Islam, Sheikh Mohammed Shariful
N1 - Publisher Copyright:
© 2022 Elsevier Ltd
PY - 2022/6
Y1 - 2022/6
N2 - This research aims to analyze the performance of state-of-the-art machine learning techniques for classifying COVID-19 from cough sounds and to identify the model(s) that consistently perform well across different cough datasets. Different performance evaluation metrics (precision, sensitivity, specificity, AUC, accuracy, etc.) make selecting the best performance model difficult. To address this issue, in this paper, we propose an ensemble-based multi-criteria decision making (MCDM) method for selecting top performance machine learning technique(s) for COVID-19 cough classification. We use four cough datasets, namely Cambridge, Coswara, Virufy, and NoCoCoDa to verify the proposed method. At first, our proposed method uses the audio features of cough samples and then applies machine learning (ML) techniques to classify them as COVID-19 or non-COVID-19. Then, we consider a multi-criteria decision-making (MCDM) method that combines ensemble technologies (i.e., soft and hard) to select the best model. In MCDM, we use the technique for order preference by similarity to ideal solution (TOPSIS) for ranking purposes, while entropy is applied to calculate evaluation criteria weights. In addition, we apply the feature reduction process through recursive feature elimination with cross-validation under different estimators. The results of our empirical evaluations show that the proposed method outperforms the state-of-the-art models. We see that when the proposed method is used for analysis using the Extra-Trees classifier, it has achieved promising results (AUC: 0.95, Precision: 1, Recall: 0.97).
AB - This research aims to analyze the performance of state-of-the-art machine learning techniques for classifying COVID-19 from cough sounds and to identify the model(s) that consistently perform well across different cough datasets. Different performance evaluation metrics (precision, sensitivity, specificity, AUC, accuracy, etc.) make selecting the best performance model difficult. To address this issue, in this paper, we propose an ensemble-based multi-criteria decision making (MCDM) method for selecting top performance machine learning technique(s) for COVID-19 cough classification. We use four cough datasets, namely Cambridge, Coswara, Virufy, and NoCoCoDa to verify the proposed method. At first, our proposed method uses the audio features of cough samples and then applies machine learning (ML) techniques to classify them as COVID-19 or non-COVID-19. Then, we consider a multi-criteria decision-making (MCDM) method that combines ensemble technologies (i.e., soft and hard) to select the best model. In MCDM, we use the technique for order preference by similarity to ideal solution (TOPSIS) for ranking purposes, while entropy is applied to calculate evaluation criteria weights. In addition, we apply the feature reduction process through recursive feature elimination with cross-validation under different estimators. The results of our empirical evaluations show that the proposed method outperforms the state-of-the-art models. We see that when the proposed method is used for analysis using the Extra-Trees classifier, it has achieved promising results (AUC: 0.95, Precision: 1, Recall: 0.97).
KW - Classification
KW - Cough
KW - COVID-19
KW - Ensemble
KW - Entropy
KW - Machine learning
KW - MCDM
KW - TOPSIS
KW - Cough/diagnosis
KW - Humans
KW - Sound
KW - COVID-19/diagnosis
KW - Machine Learning
KW - Algorithms
UR - http://www.scopus.com/inward/record.url?scp=85126562265&partnerID=8YFLogxK
UR - http://www.scopus.com/inward/citedby.url?scp=85126562265&partnerID=8YFLogxK
U2 - 10.1016/j.compbiomed.2022.105405
DO - 10.1016/j.compbiomed.2022.105405
M3 - Article
C2 - 35318171
AN - SCOPUS:85126562265
SN - 0010-4825
VL - 145
JO - Computers in Biology and Medicine
JF - Computers in Biology and Medicine
M1 - 105405
ER -